Thinking about the role of sizing in a portfolio. Some PM's size on conviction (not valuation, but quality of thesis, set up, path to getting paid), others purely on valuation (so double down if it goes against them all else unchanged), others on both. It's got me thinking 1/n

Obviously everyone who is successful over a long time in this business has both an established research process and a portfolio construction process that has worked for them. And used to work for someone who preached scaling in and out of positions as valuation evolved 2/n

And yet, it seems like all that sizing means you constantly are unwinding your successful trades as they become successful and sizing up your bad trades as the hole gets deeper and deeper. 3/n

Whereas if you never make money on high conviction trades and only on low conviction trades, you very quickly recognize your intuition as an investor is off or non-existent and can then optimize sizing, or work on that part of your process. 4/n

I guess this is a monologue about how I've started to come around to the fact that sizing based on conviction is the way to go. Maybe give myself 1 "valuation" size up, where if its gone against you can size up to the original notional dollar position. 5/n

But then exiting trades. If don't size on valuation on the way in, you don't on the way out either. Think the one exception here is if a position has done so well it now dominates the portfolio (this is in the day job where sharpe matters, in the PA, who cares just make $) 6/n

I think what this leads to is it forces me to not fall in love or get emotional about positions, especially those that have been big winners or losers. If it's time to exit, exit, don't leave cost basis or P&L gains in there for sentimental purposes to "let it ride" 7/n

And it's a good bottoms up approach towards "if the portfolio was being constructed from a blank page today, what would it be" which I think will lead to better returns and also higher quality returns 8/n

Some of this may not apply to you at all if you take the "20 hole punch card" approach, then your basically an ambush predator lying in wait for the next Amazon to cross your path, and if your good at that, portfolio construction isn't so important 9/n

But I also think this sizing would allow me to stay in an Amazon like situation or at least re-visit it with rigor and discipline if I did sell out entirely instead of saying "oh I still have some" which is a great excuse to not re-underwrite a position 10/n

Just some thoughts I've been having the last few days. If anyone has seen academic literature refuting this thought process, would love to hear about it.

It's boring and not as sexy as talking the next great idea, but this sort of stuff should be discussed more than it is 11/11